Research Methods Flashcards
All AS and A2 research methods on the AQA psychology specification are covered as part of this deck
Significance level most often used in Psychology
P<0.05
Determination that the results are significantly strong to reject the null hypothesis
Significance
The likelihood that certain events will occur
Probability
The value a test statistic must reach in order to consider the result significant
Critical value
The value of a statistic that is calculated for a particular data set (outcome of the stats test)
Calculated value
(False-positive) occurs if an investigator falsely rejects a null hypothesis that is actually true in the population
Type I error
(False-negative) occurs if an investigator fails to reject a null hypothesis that is actually false in the population
Type II error
Decreasing the probability of a type I error increases…
the probability of a type II error
How to decrease the likelihood of a Type I error
Use a smaller significance level (P<0.01)
How to decrease the likelihood of a Type II error
Increase the sample size in the study
Choose the test: Difference, nominal data, Independent Measures
Chi Squared (x2)
Choose the test: Difference, at least ordinal data, independent measures
Mann Whitney U (U)
Choose the test: difference, interval/ratio data that is normally distributed, Independent measures design
Unrelated t-test (R)
Choose the test: difference, nominal data, repeated measures
Sign test (S)
Choose the test: difference, at least ordinal data, repeated measures design
Wilcoxon t test (T)
Choose the test: difference, internal/ration data that is normally distributed, repeated measures deign
Related t-test (R)
Choose the test: Correlation, nominal data
Chi Squared (x2)
Choose the test: correlation, at least ordinal data
Spearman’s Rho (Rs)
Choose the test: Correlation, interval/ratio data that is normally distributed
Pearson’s product moment
Sign test: should the calculated value be greater than or less than for significance?
Equal to or less than
Mann Whitney U: should the calculated value be greater than or less than for significance?
Equal to or less than
Wilcoxon: should the calculated value be greater than or less than for significance?
Equal to or less than
Chi Squared: should the calculated value be greater than or less than for significance?
Equal to or greater than
Unrelated t test: should the calculated value be greater than or less than for significance?
Equal to or greater than
Related t test: should the calculated value be greater than or less than for significance?
Equal to or greater than
Spearman’s Rho: should the calculated value be greater than or less than for significance?
Equal to or greater than
Pearson’s Product moment: should the calculated value be greater than or less than for significance?
Equal to or greater than
Which inferential statistic do I have to calculate by hand?
Sign test
Qualitative data
Expressed in words, in depth-non numerical (thoughts, feelings etc).
Quantitative data
Numerical data
Primary data
Obtained first hand by the researcher for the purpose of the current study
Secondary Data
Information already collected by others or already in existence so pre-dates current project
Meta-analysis
Combining the findings from a number of studies on a particular topic for an overall conclusion
Examples of descriptive statistics
Central tendency, measures of dispersion, charts and graphs
Measures of central tendency show what
The average of a set of data (mean, median or mode)
The mean is..
found by adding all scores and dividing by the number of scores - the arithmetic average
The median is..
The central/middle value (once all data is in order)
The mode is…
The most frequently occurring value
Strength of the mean
Most representative as it uses all the data scores
Weakness of the mean
Easily distorted by outliers
Strength of the median
Unaffected by extreme scores/outliers
Weakness of the median
Less sensitive (ignores extreme values that may be important!)
Strength of the mode
easy to calculate
Weakness of the mode
Crude, not very useful, doesn’t represent all the data, might me more than one!
Measures of dispersion show..
The spread of data (range or standard deviation)
The range is
overall data spread. Biggest value minus the smallest value
Strength of the range
Easy to calcuate
Weakness of the range
Doesn’t take into account all data, extreme values only
The standard deviation is
a measure of the average deviation from the mean
How do you calculate standard deviation?
Step 1: Find the mean.
Step 2: For each data point, find the square of its distance to the mean.
Step 3: Sum the values from Step 2.
Step 4: Divide by the number of data points.
Step 5: Take the square root.
What type of graph should I use to represent the strength and direction of a correlation?
Scatter graph/Scattergram
When plotting a scatter gram which co-variable goes on the X axis?
It doesn’t matter!
What type of graph represents the frequency of each variable by the height of its spaced columns and is only used with Discrete data?
Bar Chart
When plotting a bar chart which variable goes on the X axis?
The categories (the IV)
What type of graph displays continuous data and uses the height and width of touching bars to show frequency?
Histogram
The X axis of a histogram must start at..
True 0
What does the Y axis of a histogram represent?
the frequency
A summary table would display…
Raw data
A normal distribution graph looks like…
a Bell curve, symetrical
In a normal distribution the mean, median and mode are
The same (at the highest peak of the curve)
In a normal distribution the tails of the curve never touch what?
The X axis (they are never 0)
A distribution curve that is not symmetrical is known as..
a Skewed distribution
In a positively skewed distribution the curve leans..
towards the Y axis
In a positively skewed distribution, the longer tail is on the left or right?
The right
In a negatively skewed distribution the curve leans..
away from the Y axis
In a negatively skewed distribution, the longer tail is on the left or right?
left
In a skewed distribution curve, the measures of central tendency appear in what order (starting at the top of the curve working down the tail)?
Mode, median, mean
In a normal distribution curve - what percentage of people appear within 1 standard deviation of the mean (either side)?
68%
In a normal distribution curve - what percentage of people appear within 2 standard deviations of the mean (either side)?
95%
In a normal distribution curve - what percentage of people appear within 3 standard deviations of the mean (either side)?
99.7%
A positive correlation is
As one variable increases so does the other
A negative correlation is
As one variable increases the other decreases
Zero correlation is
No relationship between variables
How can we tell the strength of a correlation
How tightly packed the points are around the line of best fit or correlation coefficient
A statistical measure of the strength and direction of a linear relationship between two variables is known as..
Correlation Coefficient
Correlation Coefficient is between..
+1 and -1
A correlation coefficient of 0.1-0.3 (positive or negative) is what strength relationship?
Weak
A correlation coefficient of 0.4-0.6 (positive or negative) is what strength relationship?
Moderate
A correlation coefficient of 0.7-0.9 (positive or negative) is what strength relationship?
Strong
A correlation coefficient of 1 (positive or negative) is what strength relationship?
A perfect correlation
DV Data which is in rank order (ranked or rated)
Ordinal data
DV data which is categorical
Nominal data
DV data which is measured in fixed units with equal distance between points on the scale or set intervals
Interval data
Which variable determined the level of data in a study?
Dependent variable
Name a qualitative method of data analysis
Content analysis or thematic anaylsis
In a content analysis what technique is performed on the data?
Coding
Content analysis converts qualitative data to
Quantitative data
Thematic analysis produces what type of data?
Qualitative
strucLab experiment is..
IV is directly manipulated under controlled setting to measure the impact on a DV
Field experiment is
IV is directly manipulated in natural settings to measure the impact on a DV
Natural experiment is
The CHANGE in an IV is natural and would have occurred without the presence of the researcher (e.g. before and after a natural disaster - no direct manipulation by a researcher) DV is measured
a Quasi experiment is
IV is already occurring (e.g. ethnicity) and so cannot be directly manipulated - the impact on DV is measured
Strength of a lab experiment
controlled conditions increase reliability and internal validity
Weakness of a lab experiment
setting isn’t natural = behaviour isn’t natural and results lack ecological validity
strength of a field experiement
Higher ecological valdity
weakness of a field experiment
lacks control and internal validity
strength of a natural experiment
high external validity, unique study opportunity